|  | Home
 | Calendar
 | Mail Lists
 | List Archives
 | Desktop SIG
 | Hardware Hacking SIG Wiki | Flickr | PicasaWeb | Video | Maps & Directions | Installfests | Keysignings Linux Cafe | Meeting Notes | Linux Links | Bling | About BLU | 
On Wed, Sep 25, 2002 at 03:43:06PM -0400, ron.peterson at yellowbank.com wrote:
> boot=/dev/sda
> image=/vmsomethingorother
>   root=/dev/md2
> 
> Otherwise, like you say, how could the system find md stuff before there
> is a kernel.
> 
> But you were sucessfully booting before...  (?)
I am successfully booting now, on a two-disk machine, with the
following lilo:
  prompt
  timeout=50
  default=linux
  boot=/dev/md6
  map=/boot/map
  install=/boot/boot.b
  message=/boot/message
  vga=0x303
  linear
  image=/boot/vmlinuz-2.4.18-10
          label=linux
          initrd=/boot/initrd-2.4.18-10.img
          read-only
          root=/dev/md5
No mention of any hda or hdc.  I don't know how it works, but it
works.
What I don't know how to do is:
 - Boot my above working machine with the first disk missing.
 - Boot my co-worker's machine with the first disk blown away by
   Debian (plugged in or not).
I think these are the same situation, for we have the same
motherboard, approximately the same hard disks, and the same 7.3
version of Red Hat.
For bootable software raid 1 to be worthwhile, it seems that figuring
out how to operate when either disk dead is key.
-kb